What a septum actually does

I have watched a lab buy 20,000 septa on price alone and then spend three weeks chasing ghost peaks, split injections, and a coring problem that turned out to be a 1 mm puncture in a 3 mm center. The septa were fine on paper. They were wrong for the method. Choosing the right septum is not a catalog exercise. It is a short engineering decision driven by temperature, needle gauge, injection count, sample chemistry, and the vial neck you actually have on the bench.

This guide is what I would hand a new buyer or a lab manager rebuilding a consumables list. It covers material chemistry, dimensions, temperature ceilings, puncture budgets, torque, and the failure modes that cost you instrument time. Where a number matters, I give the number. Where there is a tradeoff, I take a side.

A septum is a resealable elastomeric barrier that keeps a vial headspace closed to the atmosphere while allowing a needle to pass through it, sometimes dozens of times. That is the whole job. Everything else is a compromise between two competing demands: sealing and puncturing.

Sealing means low gas permeability, good compression set recovery, and chemical inertness. Puncturing means soft enough to penetrate cleanly, resilient enough to close behind the needle, and tough enough not to core. No single material does both perfectly. That is why the market is full of laminates, coatings, and pre-slit designs.

The failure modes I see most often, in rough order of frequency:

  • Coring: a plug of septum pushed into the vial, then aspirated into the needle or left to extract into the sample.
  • Ghost peaks and carryover: volatiles or extractables bleeding from the septum into the headspace.
  • Leakage after multiple injections: the puncture channel no longer seals, so headspace pressure drops and quantitative GC results drift.
  • Needle bending or septum push-through: a soft septum on a loosely crimped vial lets the needle push the whole septum into the neck.
  • Adhesion failure in laminates: the PTFE film separates from the silicone and the septum delaminates under heat.

Each of these has a specific cause and a specific fix. Most of them are preventable at the purchasing stage.

The material menu, with numbers

Almost every septum sold today is one of five constructions. Know these and you can specify 90 percent of what your lab needs.

Silicone (usually the core layer)

Silicone rubber is the workhorse. It has excellent resealing (compression set recovery above 90 percent at moderate temperature), very good puncture resilience, and a wide service range. Standard silicone septa are rated to about 200 °C continuous, with premium grades to 250 °C for short excursions. The weakness is permeability: silicone passes gases and vapors more readily than butyl or PTFE, and it can bleed low-molecular-weight siloxanes at high temperature.

For GC headspace and HPLC, silicone is almost always the core, protected by a PTFE face.

PTFE-faced silicone (the default choice)

A 0.05 mm to 0.13 mm PTFE film laminated to a silicone core is the most common septum in analytical labs. The PTFE face contacts the sample and provides near-universal chemical inertness; the silicone provides the reseal. This construction is what most people mean when they say “HPLC septum.”

Two cautions. First, the PTFE film must be oriented toward the sample. Install it backwards and you get silicone extractables in your injection. Second, cheap laminates delaminate above 180 °C or after aggressive solvent contact. If you run high-temperature GC, buy a bonded, not just laminated, product.

PTFE/silicone/PTFE sandwich

PTFE on both faces. Slightly more expensive, but it protects the silicone from solvent vapor on the headspace side and from the sample on the liquid side. I specify this for HPLC methods with aggressive mobile phases (high acetonitrile, buffered aqueous, or anything with dichloromethane) and for storage vials where the septum sits in contact with vapor for weeks.

Butyl rubber

Butyl has the lowest gas permeability of the common septum materials, which makes it the choice for headspace GC where you care about maintaining pressure and preventing oxygen ingress. Its temperature ceiling is lower, around 120 °C to 140 °C, and it is less resilient to repeated puncture. Use butyl when you inject once and care about the seal, not when you inject 50 times from the same vial.

Natural rubber and EPDM

Natural rubber is cheap, resilient, and fine for low-cost, low-temperature, single-use applications such as short-term sample storage. EPDM is a good general-purpose alternative with better chemical resistance than natural rubber and a service range to about 150 °C. Neither belongs in a high-sensitivity LC-MS method.

Matching the septum to your vial neck

The septum has to fit the closure, and the closure has to fit the vial. Get the dimensions wrong and no material choice saves you.

The two dominant neck standards in analytical labs are 11 mm (crimp and snap) and 13 mm (crimp), plus the 9 mm short-thread screw neck on 2 mL autosampler vials and the 8-425 and 10-425 screw threads. The septum diameter must match the closure, not the vial opening. A 11 mm septum in a 13 mm crimp cap will not seal, no matter how hard you crimp.

For a 2 mL autosampler vial with a 9 mm screw thread, the standard septum is a 9 mm PTFE/silicone disc, typically 1.0 mm thick. For 11 mm crimp vials, the standard is an 11 mm septum, usually 1.0 mm or 1.5 mm thick depending on the cap. The 1.5 mm septa are more forgiving of needle misalignment and give more punctures before leakage, at the cost of slightly higher needle penetration force.

Thickness matters more than people think. A 1.0 mm septum in a cap designed for 1.5 mm will not compress enough to seal. A 1.5 mm septum in a 1.0 mm cap will over-compress, deform, and fail early. Buy the thickness the cap was designed for.

If you are unsure what you have on the shelf, measure. A caliper on ten septa from the same lot should read within ±0.05 mm. If your lot varies more than that, the supplier is not controlling thickness, and your crimping torque will produce inconsistent seals.

Temperature: the number that decides the material

Temperature is the single most useful selection criterion, because it eliminates most of the catalog immediately. My working rule:

  • Below 100 °C: any of the common materials will work. Choose on cost and chemical compatibility.
  • 100 °C to 150 °C: PTFE/silicone, EPDM, or butyl. Avoid natural rubber.
  • 150 °C to 200 °C: PTFE-faced silicone only, and only bonded laminates. Check the manufacturer’s continuous rating, not the peak.
  • 200 °C to 250 °C: high-temperature silicone with a premium PTFE film. Expect to pay 2 to 3 times the standard price. Verify with a blank injection at temperature before committing a method.
  • Above 250 °C: no standard septum is reliable. Change the method or accept that the septum is a consumable with a short life.

Peak temperature ratings are marketing. Continuous ratings are engineering. If a supplier quotes only a peak, ask for the continuous rating in writing. A septum rated to 250 °C peak may have a continuous limit of 200 °C, and your GC oven sits at 220 °C for 30 minutes per run.

Puncture budget: how many injections before you change

Every septum has a finite number of clean punctures before the channel leaks or cores. This is the number I wish more buyers asked for, because it drives the true cost per injection.

From my own bench testing and supplier data, typical values for a 1.0 mm PTFE/silicone septum with a 22-gauge needle:

  • Standard PTFE/silicone: 50 to 100 punctures before leakage becomes measurable. I plan on 50.
  • Pre-slit PTFE/silicone: 100 to 150 punctures, but with a higher risk of evaporation from the slit and slightly worse seal on low-boilingsolvent samples.
  • High-resilience silicone (no PTFE): 100 to 200 punctures, but with extractables risk.
  • Butyl: 20 to 40 punctures. Do not reuse butyl septa.

Needle gauge matters as much as material. A 26-gauge needle (0.26 mm OD) makes a smaller hole and gives more punctures than a 22-gauge (0.72 mm OD). If your autosampler uses a fixed needle, size the septum to the needle, not the other way around. And never use a septum with a needle larger than the manufacturer’s rated maximum; a 16-gauge needle through a 1.0 mm septum will core it on the first injection.

Pre-slit septa are worth the extra cost when you run more than 50 injections per vial or when your autosampler needle alignment is marginal. They are not worth it for headspace GC or for volatile analytes, because the slit is a leak path.

Coring: the failure that ruins methods

Coring is a septum fragment pushed into the vial. It is the most expensive septum failure because it can plug a needle, contaminate a sample, and produce a result that looks like a real peak. I have seen a coring fragment produce a 3 percent RSD failure on a validated method and a full revalidation.

Three causes, in order of frequency:

  • Needle dullness or a burr. A needle that has been used 500 times will core a septum that a new needle passes cleanly. Replace needles on a schedule, not on failure.
  • Needle misalignment. If the autosampler needle hits the septum off-center, it pushes rather than pierces. Check alignment every time you change a vial lot or a tray.
  • Septum too soft or too thin for the needle. A 1.0 mm natural rubber septum with a 21-gauge needle will core. A 1.5 mm PTFE/silicone will not.

Pre-slit septa reduce coring dramatically because the needle follows the slit rather than tearing new material. For high-throughput labs, I consider pre-slit the default and non-slit the exception.

Chemical compatibility: what the PTFE face does and does not protect

The PTFE face is inert to almost everything, which is why it is the standard. But the PTFE is only 0.05 mm to 0.13 mm thick. It protects the sample from the silicone core; it does not protect the silicone from solvent vapor that migrates around the edge or through a puncture channel.

Practical compatibility rules I use:

  • Aqueous and buffered mobile phases: any PTFE-faced septum is fine.
  • Acetonitrile, methanol, THF: PTFE-faced silicone is fine; check for swelling of the silicone if the septum sits in vapor for more than 24 hours.
  • Dichloromethane, chloroform, hexane: use a PTFE/silicone/PTFE sandwich and change septa every run. Silicone swells in DCM.
  • Strong acids and bases: PTFE face is fine; the silicone core is not. Use a PTFE-only or PTFE-faced butyl septum if the vapor is aggressive.
  • Derivatization reagents (silylating agents, diazomethane): use a fresh PTFE-faced septum per batch and never reuse.

If you are running methods under U.S. EPA SW-846, the method specifies the sample container and preservation, not the septum brand, but the septum is part of the container. A leaking septum is a preservation failure. For volatile organics under SW-846, use a PTFE-faced silicone septum with a crimp or screw cap rated for the holding time, and document the septum lot in your QC records.

Crimp versus screw: the closure decision comes first

You cannot choose a septum without choosing a closure, and the closure determines the septum dimensions. My guidance:

  • Crimp (11 mm or 13 mm): best seal, lowest evaporation, lowest cost per vial. Requires a crimper and consistent crimping force. Use for GC, headspace, and long-term storage. Crimping torque is not a torque wrench value; it is a crimper setting. Set it so the cap deforms the septum by about 20 to 30 percent of its thickness and the vial will not rotate under moderate hand pressure. Recheck the setting every 1,000 crimps.
  • Screw thread (9 mm, 8-425, 10-425): reusable, no tool, faster for small batches. Seal quality depends on cap torque. Hand-tight plus a quarter turn is about 1.5 to 2.0 N·m for a 9 mm cap; over-torque deforms the septum and causes coring. Use a torque wrench if you run validated methods.
  • Snap caps: convenient, but the seal is less consistent than crimp. Fine for short holding times, not for volatile analytes.

The vial and the septum are a system. A premium septum on a poorly molded vial neck will leak. Check neck dimensions against the GPI 9 mm and GPI 11 mm thread standards, and buy vials and closures from the same supplier when you can, so the tolerances are matched.

A decision checklist with numbers

Here is the checklist I actually use when specifying septa for a new method. Work top to bottom; each answer narrows the catalog.

  • Temperature: maximum continuous oven or autosampler temperature. If above 200 °C, specify high-temperature silicone with a bonded PTFE film and plan to change every 50 injections.
  • Needle gauge and type: fixed or removable, 22G or 26G. Match septum thickness to needle: 1.5 mm for 21G and larger, 1.0 mm for 22G to 26G.
  • Injections per vial: under 10, any PTFE-faced silicone; 10 to 50, standard PTFE/silicone; 50 to 150, pre-slit PTFE/silicone; over 150, change vials, not just septa.
  • Sample chemistry: aqueous or buffered, standard PTFE-faced silicone; halogenated solvents, PTFE/silicone/PTFE sandwich and single use; derivatization, single use.
  • Headspace vs liquid injection: headspace, butyl or PTFE-faced silicone with a low-permeability core, no pre-slit; liquid, pre-slit is acceptable.
  • Closure: crimp for long holding times and volatiles; screw for convenience and small batches; torque 1.5 to 2.0 N·m for 9 mm screw caps.
  • Dimensions: 9 mm septum for 9 mm screw vials, 11 mm for 11 mm crimp, 13 mm for 13 mm crimp; thickness 1.0 mm or 1.5 mm per cap design; caliper check ±0.05 mm.
  • Lot documentation: ask for a certificate of analysis with durometer, thickness, and extractables data. If the supplier cannot provide it, the price is not a bargain.

What to buy, in plain terms

For a general HPLC lab running reversed-phase methods at ambient to 40 °C with a 22-gauge needle and fewer than 50 injections per vial: 9 mm or 11 mm PTFE/silicone septum, 1.0 mm thick, non-pre-slit, from a supplier who provides a CoA. Buy in lots of 1,000 and rotate stock. This covers most of what you will do.

For a GC lab running headspace at 80 °C to 150 °C: 11 mm PTFE-faced silicone, 1.5 mm thick, butyl core if you need the lowest permeability, non-pre-slit. Change every 20 to 40 injections.

For a high-throughput autosampler running 100 or more injections per vial: pre-slit PTFE/silicone, 1.0 mm, 9 mm screw, and replace needles on a schedule. The pre-slit pays for itself in coring reduction.

For high-temperature GC above 200 °C: bonded PTFE/silicone, 1.5 mm, high-temperature grade, and run a blank injection at temperature to confirm no bleed before you commit the method.

Operational precautions that save money

  • Store septa in the original sealed bag, away from solvent vapors and sunlight. Silicone absorbs volatiles from the air; a septum stored open in a lab with a solvent cabinet will show extractables.
  • Do not touch the PTFE face with bare fingers. Oils and salts transfer. Use forceps.
  • Orient the PTFE face toward the sample. Mark the bag or the tray so the operator cannot get it wrong at 2 a.m.
  • Check needle alignment after every tray change. A 0.5 mm misalignment is enough to start coring.
  • Replace needles every 500 to 1,000 injections, or sooner if you see coring. A dull needle costs more in failed runs than a box of needles costs.
  • Run a system suitability check after every septum lot change. A new lot is a new material, even from the same supplier.

Troubleshooting checklist

When results drift, work this list before you blame the column.

  • Ghost peaks in blank injections: septum bleed. Check temperature rating and PTFE orientation; try a different lot.
  • Decreasing peak area over a run: leakage at the puncture channel. Reduce injections per vial or switch to pre-slit.
  • RSD above 2 percent on a method that used to pass: coring or a partially blocked needle. Inspect the septum for a missing center and the needle for a burr.
  • Sample loss in storage: evaporation through the septum. Check cap torque and septum thickness; a 1.0 mm septum on a cap designed for 1.5 mm is a leak.
  • Septum pushed into the vial: over-soft septum or under-crimped cap. Increase crimp setting or use a 1.5 mm septum.
  • PTFE film separated from the silicone: delamination from heat or solvent. Switch to a bonded laminate and lower the temperature.

Where to source and what to ask

Buy septa from a supplier who publishes continuous temperature ratings, durometer (typically 50 to 60 Shore A for analytical silicone), thickness tolerance, and extractables data. If those numbers are not on the datasheet, ask for them in writing. A supplier who cannot answer a durometer question cannot control a lot.

For method development and cross-checking, the technical libraries at Restek and the application notes from Agilent and Thermo Fisher are useful for understanding how septum bleed and leakage show up in real chromatograms. The vial septa catalog at vialsepta.com lists the standard 9 mm, 11 mm, and 13 mm constructions with their temperature ratings, and the HPLC vial pages cover the matching closures and dimensions.

My recommendation for most labs: standardize on one PTFE/silicone septum in one thickness and one diameter per vial type, buy it from a supplier with a CoA, and put the lot number on every method record. The savings from switching suppliers to save 15 percent on a box of 1,000 septa disappear the first time a coring fragment costs you a revalidation. Pay for the controlled part, change it on a schedule, and spend your time on the chemistry.

References & further reading

Related resources from our lab consumables network:

External references and standards: